The content of the invention
Due to scheme human factor of the prior art influence it is larger, the present invention provide one kind be affected by human factors compared with
It is small, the pliability method of testing and system of the high automation of test accuracy.
A kind of pliability method of testing of the present invention, including step:S1, in test zone test equalizing bed and zinc coat steel plate is set,
It is initial to shoot test zone image, make the position of the test equalizing bed and zinc coat steel plate in the test zone image;S2、
Test zone image is shot, judges test on equalizing bed whether have human body image described in test zone image, if so, then performing step
Rapid S3;If it is not, then repeating step S2;S3, continuation shoot test zone image, and whether the hand for monitoring human body image is relative
It is static after zinc coat steel plate persistent movement, and the time of static maintenance reach a default time threshold, if so, then performing step S4;
If it is not, then repeating step S3;The leg of human body image is surrounded with test equalizing bed when hand in S4, calculation procedure S3 is static
Region area, judge the region area whether be more than one setting area threshold, if then prompt action is nonstandard;If
It is no, then perform step S5;S5, calculating analyze the relative position relation of the now hand of human body image and zinc coat steel plate, are tested
As a result.
As a kind of preferred scheme, the test equalizing bed and zinc coat steel plate are made in step S1 in the test zone image
In position comprise the following steps:S11, a mark is set on zinc coat steel plate;S12, shooting test zone image, detect zero
Mark on the plate of position, so that it is determined that going out position of the zinc coat steel plate in test zone image;S13, according to test equalizing bed and zero-bit
The position relationship of plate makes position of the test equalizing bed in the test zone image;
Hand is also defined in step S1 there is domain and leg and domain occur, comprise the following steps:S14, with test zone
On the basis of zinc coat steel plate in image, determine that domain occurs in a hand above zinc coat steel plate;S15, with zero in test zone image
On the basis of position plate and test equalizing bed, determined in zero-bit board test towards between tester side and test equalizing bed upper surface
There is domain in one leg;In each step, it will appear in the human body image part that domain occurs in the hand and be defined as hand, will appear in
The human body image part that domain occurs in the leg is defined as leg.
As a kind of preferred scheme, when shooting test zone image, by the background of test zone be set to reflective fabric and
The human body image provided during shooting in light source irradiation test zone, each step need to carry out binary conversion treatment;Also made in step S1
Human body pixel threshold value, including:People keeps correct leg gesture, and sole withstands zinc coat steel plate and is sitting on test equalizing bed, and shooting is obtained
Take test zone image;The pixel value of the leg image after binaryzation is counted, the minimum that can interpolate that during for sitting human body is made
Pixel value, as human body pixel threshold value;It is judged as having the pixel value of human body image section need to be more than the human body picture in step S2
Plain threshold value.
The area surface that human body image leg is surrounded with test equalizing bed is calculated as a kind of preferred scheme, in step S4
Product, including:The midpoint of each row of the leg image after binaryzation is calculated, it is bent that all row midpoints are in turn connected to form into leg
Line, calculates the area that the legs curve is surrounded with test equalizing bed.
As a kind of preferred scheme, the unit pixel and reality in test zone image are also precomputed in the step S1
Proportionate relationship between the length value of border;The step S5 includes:Hand images after S51, line by line search binaryzation, obtain finger tip
Position;Whether S52, judgement fingertip location are located at side of the zinc coat steel plate towards tester, if so, then test result is zero;
If it is not, then performing step S53;S53, maximum normal distance between finger tip image and zinc coat steel plate image is measured, then according to survey
The proportionate relationship between unit pixel and actual (tube) length angle value in examination area image calculates actual distance value, calculates test result,
Obtain test result.
A kind of pliability test system of the present invention, including:Test zone, includes test equalizing bed and positioned at test equalizing bed
The zinc coat steel plate of one end;Taking module, for shooting test zone image;Locating module, for make it is described test equalizing bed and
Position of the zinc coat steel plate in the test zone image;Human bioequivalence module, for receiving test zone image and described fixed
The position position made of module, and calculate and judge whether there is human body image in test equalizing bed described in test zone image;If then
Determination module, which is completed, to action sends enabling signal;Action completes determination module, for receiving the transmission of human bioequivalence module
Enabling signal when start and receive test zone image, monitoring human hands it is whether static with respect to after zinc coat steel plate persistent movement,
And the time of static maintenance reaches a default time threshold, if then acts of determination completes and sent to posture determination module to open
Dynamic signal;Posture determination module, for starting when receiving action and completing the enabling signal of determination module transmission and receiving survey
Area image is tried, the region area that human leg surrounds with test equalizing bed is calculated, judges whether the region area is more than one
The area threshold of setting, if otherwise judging posture standard, and sends enabling signal to test result statistical module;And test knot
Fruit statistical module, for starting when receiving the enabling signal of pose discrimination module transmission and receiving test zone image, is counted
The relative position relation of human hands and zinc coat steel plate is calculated, test result is obtained.
As a kind of preferred scheme, a mark is posted on the zinc coat steel plate, the locating module includes:Origin reference location list
Member, the test zone image for receiving taking module shooting, detects the mark on zinc coat steel plate, so that it is determined that going out zinc coat steel plate
Position in the picture, and the test equalizing bed is made in the test zone according to the position relationship of test equalizing bed and zinc coat steel plate
Position in image;Zone location unit, zinc coat steel plate and test have been set for receive that origin reference location unit sends
The test zone image of equalizing bed position, determines that the human body image that a hand domain occurs so that the region to be occurred is true above zinc coat steel plate
It is set to hand, the human figure that a leg domain occurs so that the region to be occurred is determined between zinc coat steel plate and test equalizing bed upper surface
As being defined as leg.
As a kind of preferred scheme, the pliability test system also includes:Reflective fabric, for being set to test zone
Testing background;Light source, for when taking module shoots test zone image, irradiating the test zone;Binarization block, is used
Test zone image after the reception taking module shoots test zone image and finishes binary conversion treatment issues described
Locating module, the human bioequivalence module, the action complete determination module and the posture determination module;The shooting mould
Block is additionally operable to shooting tester and is sitting on test equalizing bed before the test begins, when sole withstands zinc coat steel plate and keeps correct body position
Human body image, to obtain correct body position leg image;The locating module also includes a pixel threshold setup unit, for testing
The correct body position leg image after binaryzation is received before starting, and is calculated accordingly and is made and can interpolate that leg during for sitting human body
There is the pixel threshold in domain in portion, and sends human bioequivalence module to, and human bioequivalence module the pixel value in domain occurs measuring leg
It is judged as someone during more than the human body pixel threshold value.
As a kind of preferred scheme, the posture determination module includes:Legs curve formation unit, for receiving binaryzation
Leg image after module binary conversion treatment, calculates the midpoint of each row of leg image, all row midpoints successively line is formed
Legs curve;Areal calculation unit, for receiving the legs curve that sends of legs curve formation unit, calculate the curve with
The area that test equalizing bed is surrounded is the region area that leg is surrounded with test equalizing bed.
As a kind of preferred scheme, the locating module includes a lenth ratio determining unit again, for starting in test
Test zone image after preceding reception binaryzation, calculates the ratio between the unit pixel in test zone image and actual (tube) length angle value
Example relation simultaneously sends the test result statistical module to;The test result statistical module includes:Fingertip location determining unit,
For searching for the hand images after binaryzation line by line, fingertip location is obtained;Score and just sentence unit, for receiving the survey after binaryzation
The fingertip location that examination area image and fingertip location determining unit are determined, judges whether fingertip location is located at the zinc coat steel plate
Towards the side of tester, if so, then test result fraction is directly zero, otherwise enabling signal is sent to score value computing unit;
Score value computing unit, for the test zone after the proportionate relationship and binaryzation between recruiting unit's pixel and actual (tube) length angle value
Image and fingertip location, measure the maximum normal distance between finger tip and zinc coat steel plate image, and finger is calculated according to the proportionate relationship
The actual distance value of point and zinc coat steel plate, calculates test result, obtains test result.
A kind of pliability method of testing of the present invention and system, pliability test is carried out using image recognition, and automatic identification is surveyed
Examination process posture and record information, not only substantially reduce influence of the human factor to test result, and test accuracy is high.
Embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, it is right below in conjunction with drawings and Examples
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used to limit the present invention.
The pliability method of testing and system of the present invention can be used in the pliability test of national physique detection.
Fig. 1, Fig. 2 are refer to, pliability test system of the present invention includes:Test zone 1, taking module 2, locating module 3,
Human bioequivalence module 4, action complete determination module 5, posture determination module 6, test result statistical module 7 and binarization block
8。
Test zone 1 taking module 2 shooting within sweep of the eye.Test zone 1 includes test equalizing bed 101 and zinc coat steel plate
102, zinc coat steel plate 102 is arranged on one end of test equalizing bed 101 perpendicular to test equalizing bed 101.It can be conventional measurement to test equalizing bed 101
The equalizing bed used in examination, with higher hardness, to reduce due to the shadow that the elasticity of test equalizing bed 101 is caused to test result
Ring.
Referring to Fig. 3, during test, tester is sitting on test equalizing bed 101, and both feet prop up zinc coat steel plate 102, the both hands centre of the palm to
Under, both arms close up flattened, high low hand, and finger tip is crossed the maximum distance of zinc coat steel plate 102 and sentenced as body pliability using in both hands
Disconnected standard.
Taking module 2 is used to shooting the image of test zone 1, and the camera lens of taking module 2 is towards test zone 1, specifically
Say, towards test equalizing bed 101 and zinc coat steel plate 102, people's body side surface is just seated to taking module 2 on test equalizing bed 101.In this implementation
In example, taking module 2 is thermal camera.And in order to quickly make the position of test equalizing bed 101 and zinc coat steel plate 102 in the picture
Put, and convenient follow-up progress target detection, the background of test zone 1 is provided with one piece of reflective fabric 103, and provides infrared light supply
(not shown) irradiates test zone 1, and is shot using thermal camera.The shooting environmental, through whole body examination
Journey.
Further, in order that the image that taking module 2 is photographed can accurately detect zinc coat steel plate 102, in zinc coat steel plate
A mark (not shown) is posted on 102, the mark is the material for being not easy to reflection light, such as black material.Due to
Background is that the index point or human body of black material in reflective fabric, image are weak compared with the ability of reflective fabric reflected light, therefore the figure clapped
The brightness of image of mark or human body parts will be less than the brightness values of other parts in image as in, mark or human body parts with
The bright environment of background presents a contrast.
The test zone image photographed for taking module 2, system provides binarization block 8 and first carried out at binaryzation
Reason, the image after processing is sent to locating module 3, human bioequivalence module 4, action completes determination module 5 and posture judges mould
Block 6 so that the image that each module is analyzed is the image after binary conversion treatment.
Test starts advance rower and determined, and first makes test equalizing bed 101 and zinc coat steel plate 102 by locating module 3 and is clapping
Position in test zone image.Specifically, locating module 3 includes an origin reference location unit.The survey shot for taking module 2
Area image is tried, is made via binarization block 8 after binary conversion treatment, origin reference location unit first detects zinc coat steel plate 102 in image
On mark.Referring to Fig. 4, origin reference location unit is to image using processing methods such as Hough transforms, and detection is obtained in image
Black line, the black line is the mark being provided on zinc coat steel plate 102.Due to mark on zinc coat steel plate 102 position, it is known that because
This, according to the position of the mark detected, you can make the position for surveying zinc coat steel plate 102 in test zone image, then, root
According to the position relationship of test equalizing bed 101 and zinc coat steel plate 102, position of the test equalizing bed 101 in test zone image is determined.
As a further optimization of the scheme , locating module 33 also includes a zone location unit.Test equalizing bed is being determined
101 with behind position of the zinc coat steel plate 102 in test zone image, zone location unit determines that hand goes out in the top of zinc coat steel plate 102
Show domain the human body image that the region occurs is defined as into hand;Leg is determined between zinc coat steel plate 102 and the upper surface of equalizing bed 101
There is domain the human body image that the region occurs is defined as into leg in portion.There is domain and leg appearance in hand in test zone image
Being determined in advance for domain, facilitates in test process, human hands and leg need to be gestured analysis when, directly extract the hand and leg
There is the image in domain in portion, so as to reduce hunting zone.
In addition, locating module 3 also includes a pixel threshold setup unit, for setting a pixel threshold before testing, it is
Judge to test in image during follow-up human testing and whether people's offer foundation is provided on equalizing bed 101.The setting process of the pixel threshold
For:Referring to Fig. 3, tester is sitting on test equalizing bed 101, and both feet prop up zinc coat steel plate 102, and taking module 2 shoots test zone figure
Picture.Obtained image is shot after binary conversion treatment is made via binarization block 8, locating module 3 counts the pixel value in leg domain,
The value is pixel threshold.Referring to Fig. 5, after making binary conversion treatment to the test zone image photographed, the right in image
Black portions are to be sitting in the leg of the test upper tester of equalizing bed 101, count the partial pixel value, that is, obtain pixel threshold.
After the demarcation that locating module 3 completes before test, the human body attitude detection of body examination process is carried out.
Test starts, and taking module 2 shoots test zone image, after being handled via binarization block 8, human bioequivalence module
4 first to detect whether that someone is seated at equalizing bed 101 first-class to be tested.Specifically, human bioequivalence module 4 calculates leg in test zone image
There is the pixel value of domain part in portion, judges whether the value is more than the pixel threshold that positioning unit 3 is set, if it is not, then judging do not have
People, then continue to the test zone image that taking module 2 is photographed, and carries out same analysis;Conversely, then judging someone.
Judge someone be seated at equalizing bed 101 it is first-class to be tested when, human bioequivalence module 4 to action complete determination module 5 send enabling signal.
After the enabling signal that human bioequivalence module 4 is sent is received, action, which completes the monitoring hand of determination module 5, to be occurred in domain
Hand it is whether static with respect to after the persistent movement of zinc coat steel plate 102, and the time of static maintenance reach a default time threshold, if
It is that then discriminating test action is completed, and enabling signal is sent to pose discrimination module 6, if otherwise explanation action is not completed, action
Module 5 is completed to continue to monitor.
After the enabling signal that action completion determination module 5 is sent is received, people in the detection and analysis image of pose discrimination module 6
Posture whether standard, the only action under standard gestures can just be included in achievement.Further, pose discrimination module 6 includes one
Legs curve formation unit and an area computing unit.After the enabling signal that action completion determination module 5 is sent is received, leg
Portion's curve formation unit receives leg in test zone image, analysis image and the image of domain part occurs, calculates leg image every
All row midpoints are in turn connected to form legs curve by the midpoint of one row;Areal calculation unit calculates the legs curve and test
The area that equalizing bed is surrounded is the region area that leg is surrounded with test equalizing bed, i.e. the legs curve is surrounded with equalizing bed 101
Area.Then, pose discrimination module 6 judge the region area whether be more than one setting area threshold (area threshold, can
It is manually set with the Stringency according to test), if otherwise showing, leg is not bent or degree of crook is in allowed band
It is interior, judge posture standard, enabling signal is sent to test result statistical module 7;Conversely, then illustrating that posture is nonstandard, it can use
The mode that voice and picture are shown reminds tester to keep correct set.
When detecting tester's posture standard, test result statistical module 7 is used to do last examination result.The test
As a result statistical module 7 just sentences unit and a score value computing unit including a fingertip location determining unit, a score.Receiving appearance
After the enabling signal that gesture determination module 6 is sent, there is domain part to hand in test zone image in fingertip location determining unit
Hand images are searched for line by line, obtain fingertip location.Then, score and just sentence whether unit judges fingertip location is located at zinc coat steel plate
102 towards the side of tester, if so, then test result is directly zero, if it is not, then score value computing unit measures finger tip in image
Maximum normal distance between zinc coat steel plate 102, calculates test result, provides test result.
Further, in order that the tester's finger tip detected is more directly perceived from a distance from zinc coat steel plate and be easy at score
Reason, locating module 3 also includes a lenth ratio determining unit, and it receives the test section after binary conversion treatment before the test begins
Area image, calculates the proportionate relationship between the unit pixel in test zone image and actual (tube) length angle value.So, in test result
, can bonding unit pixel and reality when the maximum normal distance between finger tip and zinc coat steel plate 102 in image is measured in the calculating of statistical module 7
Proportionate relationship between the length value of border, calculates test result, provides test result.
Compared with prior art, pliability test system of the present invention, pliability test is carried out using image recognition, automatic to know
Other test process posture and record information, not only substantially reduce influence of the human factor to test result, and test accuracy
It is high.
Meanwhile, present invention also offers a kind of pliability method of testing.Referring to Fig. 6, the method comprising the steps of:
S1, in test zone test equalizing bed and zinc coat steel plate is set, it is initial to shoot test zone image, make the testing cushion
The position of bed and zinc coat steel plate in the test zone image;
Referring to Fig. 2, test zone 100 includes test equalizing bed 101 and zinc coat steel plate 102.Before step-on testing is carried out, enter in advance
Rower is determined.The content of demarcation includes:Make the position of test equalizing bed, zinc coat steel plate in the test zone image clapped.
In the present embodiment, the position of equalizing bed and zinc coat steel plate in the picture is tested in order to quickly make, and it is convenient follow-up
Target detection is carried out, the background of test zone is set to reflective fabric by us, and provides infrared light supply irradiation test zone and profit
Shot with thermal camera.The shooting environmental, through whole body examination process.Meanwhile, on zinc coat steel plate, in such as testboard
Between, a mark is set, and the mark is the material for being not easy to reflection light, such as black material.After setting completed, infrared light supply
Test zone is irradiated, video camera shoots test zone image.Because background is reflective fabric, in image the mark of black material or
Human body is weak compared with the ability of reflective fabric reflected light, therefore the brightness of image of mark or human body parts will be less than figure in the image clapped
The environment that the image of the brightness value of other parts as in, mark or human body parts becomes clear with background presents a contrast.Referring to Fig. 4,
After the test zone image by being obtained to shooting makees binary conversion treatment, using processing methods such as Hough transforms, figure is detected
The black line as in, the black line is the mark being provided on zinc coat steel plate.Due to position of the mark on testboard, it is known that because
This, according to the position of the mark detected, you can make the position for surveying zinc coat steel plate 102 in test zone image, then, root
According to test equalizing bed and the position relationship of zinc coat steel plate, position of the test equalizing bed in test zone image is determined.
Further, will be in zinc coat steel plate after test equalizing bed and position of the zinc coat steel plate in test zone image is determined
Top determines that hand domain occurs the human body image that the region occurs is defined as into hand;Between zinc coat steel plate and equalizing bed upper surface
Determine that leg domain occurs the human body image that the region occurs is defined as into leg.In test zone image hand occur domain and
There is being determined in advance for domain in leg, facilitates in test process, human hands and leg need to be gestured analysis when, directly extract this
There is the image in domain in hand and leg, so as to reduce hunting zone.
In addition, the demarcation before test also includes the setting to pixel threshold, it is subsequently to make to judge in image during human testing
Test on equalizing bed whether someone provides foundation.During setting, tester is sitting on test equalizing bed, and both feet prop up zinc coat steel plate, is shot and is surveyed
Try area image.Obtained image is shot after binary conversion treatment is made, the pixel value in leg domain is counted, the value is pixel threshold
Value.Referring to Fig. 5, after making binary conversion treatment to the test zone image photographed, the right black portions are to sit in image
In the leg of test equalizing bed upper tester, the partial pixel value is counted, that is, obtains pixel threshold.
S2, shooting test zone image, judge test on equalizing bed whether have human body image described in test zone image, if
It is then to perform step S3;If it is not, then repeating step S2
Complete after the demarcation before test, carry out the human testing of body examination process.First, test zone image is shot, to figure
The image that domain part occurs in leg as in makees binary conversion treatment, pixel value in the image after statistical disposition, if the value is more than pixel
Threshold value, then judge someone be seated at test equalizing bed it is first-class to be tested, continue executing with step S3;If being less than, on discriminating test equalizing bed
Nobody, continue shoot test zone image, carry out image in test equalizing bed on whether the judgement of someone.
S3, continuation shoot test zone image, and whether the hand for monitoring human body image is quiet with respect to after zinc coat steel plate persistent movement
Only, and time of static maintenance reaches a default time threshold, if so, then performing step S4;If it is not, then repeating step
S3。
Judging to start at control zone after someone, continuing to shoot test zone image, in the test zone image to acquisition
The image of middle testboard top position is made after binary conversion treatment, whether monitors the hand of human body image with respect to zinc coat steel plate persistent movement
It is static afterwards, and the time of static maintenance reach a default time threshold, if so, then performing step S4;If it is not, then continuing executing with
Step S3, makees same analysis.In the step, it is necessary to carry out the static monitoring held time be because pliability test in,
If human body is to extend to a certain extreme position moment but can not maintain just to retract immediately in the position, then by the survey of the position
The result that test result is tested as the human body pliability is inaccurate, should be excluded this kind of inaccurate situation.
The area surface that the leg of human body image is surrounded with test equalizing bed when hand in S4, calculation procedure S3 is static
Product, judges whether the region area is more than the area threshold of a setting, if then prompt action is nonstandard;If it is not, then performing step
Rapid S5;
Detecting that tester's hand is static with respect to after zinc coat steel plate persistent movement, and static maintenance time reach it is default
After time threshold, continue to analyze people in image posture whether standard, the only action under standard gestures can just be included in achievement.Appearance
Whether the judging basis of standard are state, calculate the region area that the leg of human body image is surrounded with test equalizing bed, judging should
Whether region area is more than an area threshold set, and (area threshold can artificially be set according to the Stringency of test
It is fixed), if then by the way of voice and picture are shown posture can be pointed out inaccurate, and remind tester to keep correct body position;
If it is not, then showing that leg is not bent or degree of crook is in allowed band, posture standard is judged.The leg of human body image and survey
The region area calculating process that examination equalizing bed surrounds is to take leg in the test zone image for made binary conversion treatment domain occur
Partial image, calculates the midpoint of each row of leg image, all row midpoints is in turn connected to form into legs curve, and the leg is bent
The area that line is surrounded with test equalizing bed is the region area that leg is surrounded with test equalizing bed.
When the posture for testing and analyzing out people is standard, then step S5 is continued executing with, start to calculate test result.
S5, calculating analyze the relative position relation of the now hand of human body image and zinc coat steel plate, obtain test result.
In test result step is calculated, hand in the test zone image for made binary conversion treatment is taken domain part occur
Hand images are searched for by image line by line, obtain fingertip location, then judge whether the fingertip location is located at zinc coat steel plate direction
The side of tester, if so, then test result is zero;If it is not, then detecting that the maximum in image between finger tip and zinc coat steel plate is hung down
Straight distance, then according to the maximum normal distance, calculates test result.In order that the tester's finger tip detected is from zinc coat steel plate
Distance is more directly perceived and is easy to score to handle, it is necessary to having made the test zone after binary conversion treatment in demarcation before testing
Image, calculates the proportionate relationship between the unit pixel in image and actual (tube) length angle value.So, the finger tip in image is measured in calculating
During maximum normal distance between zinc coat steel plate, proportionate relationship that can be between bonding unit pixel and actual (tube) length angle value calculates survey
Fraction is tried, test result is provided.
Compared with prior art, pliability method of testing of the present invention, pliability test is carried out using image recognition, automatic to know
Other test process posture and record information, not only substantially reduce influence of the human factor to test result, and test accuracy
It is high.
For the person of ordinary skill of the art, without departing from the inventive concept of the premise, if can also make
Dry modification and improvement, these belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be with appended right
It is required that being defined.